1904.] Combining Properties of Serum-Complements, etc. 5 



only determine how much C is obtainable, and this amount will be 

 in excess of the free C, because part of it is got by dissociation. 



As is well known, Ehrlich explained the analogous phenomenon in the 

 case of diphtheria toxin and anti-toxin by supposing that in the " toxin " 

 there is a mixture of toxin in the proper sense and degenerated toxin or 

 toxoid, the latter having little toxic action and also, as a rule, a feebler 

 combining affinity. Bordet,* on the other hand, explains the phenomenon 

 in the case of complement and anti-complement by the hypothesis that 

 an anti-complement molecule has no fixed combining relationship, but 

 may become united to one, two, three, etc., complement molecules, so 

 that we have AC 1 , AC 2 , AC 3 combinations. Madsen, Dreyer, and 

 Arrhenius,t from a number of very important researches on tetanolysin 

 and diphtheria toxin, come to the conclusion that the union of toxin 

 and anti-toxin follows the chemical law of Guldberg and Waage for 

 substances whose combining affinity is somewhat feeble. In other 

 words, the combination belongs to the category of reversible actions, 

 the amount of the two substances in combination having always a 

 definite relationship to the amount of each of the two substances free. 

 We have not worked out with sufficient fullness the behaviour of C 

 md anti-C to make a definite statement on this point ; we may, 

 lowever, state that the results which we have got present a close 



mlogy to those of the Danish observers. It should be noted, 

 lowever, that the nature of the union of toxin or complement with its 

 mti-substance is one matter, the question as to the existence of toxoids 



id complementoids is another. 



(b) On the Specific Affinities of the Haptophore Group. There is 

 mother way in which the combinations of C with anti-C and R + IB 



lolecules respectively may be compared, namely, as regards specific 

 iffinities. 



1. Using the anti-C to guinea-pig's C obtained from the rabbit, 

 we find in confirmation of Ehrlich and Morgenroth that this anti-C 

 has some effect on rabbit's C, though it is very slight. In a given 

 case, for example, we found that the addition of O'l c.c. anti-C to the 

 M.H.D. of rabbit's C scarcely diminishes the amount of lysis which 

 follows when the test amount of red corpuscles treated with IB is 

 added ; there is merely a slight deposit in the tube as compared with 

 the control, and 0'6 of a M.H.D. of C has to be added before this 

 disappears, i.e., 1-6 M.H.D. treated with (H c.c. anti-C produces quite 

 complete haemolysis. With guinea-pig's complement O'l c.c. anti-C 

 with a little more than the M.H.D. of C prevents lysis almost com- 

 pletely, and three additional doses of C must be added to the mixture 

 before complete lysis occurs when the test is made in the same way 



* Bordet, ' Annales de 1'Inst. Pasteur,' vol. 17, 1903, p. 161. 

 t Vide ' Festskrif t ved Indvielsen af Statens Serum Institut,' Copenhagen, 1902 ; 

 also Madsen, ' Centralbl. f. Bakteriol.,' vol. 34, p. 630. 



